Abstract
Regression assumes that one can fit real data into a (simple) formula to obtain an accurate prediction of future data. Classification trees have their rules set without the use of a formula. Many times, they are set by law, industry standards, or regulations. The chapter also explains entropy; it is the measurement of information content of a message. Furthermore, one can get several free tree packages in UNIX, Windows, and Macintosh versions from the statistics department at the University of Wisconsin. QUEST stands for quick, unbiased, and efficient statistical tree. It is similar to the CART algorithm given in "Classification and Regression Trees by Leo Breiman. QUEST is generally faster than CART or chi-squared automatic interaction detector (CHAID), but it needs more main storage. This is fine for smaller data sets, but not so good for larger data sets. Generalized, unbiased, interaction detection and estimation (GUIDE) is a multipurpose machine learning algorithm for constructing regression trees. It has options for adjusting how the splits are made and how missing values are handled.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
